Thanks for the suggestions guys.  I'm still struggling with a couple
of things.  How do I give the user a link to expose the file from
extra_files_path as a history item and how would I set metadata like
file format and db on the extra files?  Things like PDF and text work
fine, but I'd really like to be able to have BED and WIG files in the
history with the UCSC genome browser link.

Thanks again!

On Mon, Oct 31, 2011 at 2:05 PM, Ross <ross.laza...@gmail.com> wrote:
> Hi, Jesse,
> IMHO, the Html datatype is perfect for this. By definition, the
> composite datatype can have any number of files (of any type!) in the
> files_path, but of course, your tool or wrapper will need to generate
> the legal html contents of that html page but that can be done when
> the files_path already has the outputs - the paths
>
> Please take a look at
> http://lists.bx.psu.edu/pipermail/galaxy-dev/2010-September/003311.html
> and feel free to ask here if anything needs clarification.
>
>
> On Mon, Oct 31, 2011 at 2:50 PM, Jesse Erdmann <jerdm...@umn.edu> wrote:
>> I have an in-house Perl tool that generates a double digit number of
>> outputs that we'd like the user to have access to, but don't
>> necessarily need to be in the history.  A complicating factor is that
>> the exact number is determined at run time as user input can add any
>> number of outputs that can't be determined in the scope of the tools'
>> XML processing.
>>
>> Ideally, the end result would be that executing the tool would produce
>> a single output in the history with the other files either being
>> hidden or in the extra_files path.  The visible file should be an HTML
>> file that can be used to describe and link to the other outputs as
>> well as providing the option to either unhide or add the hidden
>> outputs to the history for use in subsequent analysis.
>>
>> I've looked at composite datatypes, but as far as I can tell the
>> documentation and examples don't cover how to produce a composite
>> datatype with a variable number of outputs.  I briefly tried to use
>> the deprecated <code> tag to iterate over the results with the other
>> outputs added as hidden extra files to the index after execution, but
>> wasn't really excited about using a deprecated feature and couldn't
>> get it working quickly, regardless.  Any other suggestions?
>>
>> --
>> Jesse Erdmann
>> Bioinformatics Analyst
>> Masonic Cancer Center
>> University of Minnesota
>> jerdm...@umn.edu
>> 612-626-3123
>> ___________________________________________________________
>> Please keep all replies on the list by using "reply all"
>> in your mail client.  To manage your subscriptions to this
>> and other Galaxy lists, please use the interface at:
>>
>>  http://lists.bx.psu.edu/
>>
>
>
>
> --
> Ross Lazarus MBBS MPH;
> Associate Professor, Harvard Medical School;
> Head, Medical Bioinformatics, BakerIDI; Tel: +61 385321444;
>



-- 
Jesse Erdmann
Bioinformatics Analyst
Masonic Cancer Center
University of Minnesota
jerdm...@umn.edu
612-626-3123

___________________________________________________________
Please keep all replies on the list by using "reply all"
in your mail client.  To manage your subscriptions to this
and other Galaxy lists, please use the interface at:

  http://lists.bx.psu.edu/

Reply via email to